Banca de DEFESA: DIOGO HOFMAM

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
DISCENTE : DIOGO HOFMAM
DATA : 14/03/2022
HORA: 10:00
LOCAL: Defesa Remota
TÍTULO:

DEVELOPMENT OF A LOW COST CHEMICAL COMPOSITION APPLIED TO AUSTEMPERED DUCTILE CAST IRON THROUGH THE THE AID OF NEURAL NETWORKS

PALAVRAS-CHAVES:

Austempered Ductile Iron; Mechanical properties; Artificial neural networks; Cost reduction.

PÁGINAS: 47
GRANDE ÁREA: Engenharias
ÁREA: Engenharia de Materiais e Metalúrgica
SUBÁREA: Metalurgia de Transformação
ESPECIALIDADE: Fundição
RESUMO:

This work presents the development of a low-cost chemical composition for the production of an austempered nodular cast iron (ADI) that meets the ASTM A897/897M - 2016 Grade 2 1050/750/07 standard with the aid of artificial neural networks. Through extensive analysis of chemical compositions and mechanical properties found in the literature, information was compiled to insert into neural networks, seeking an optimization in the relationship between chemical composition, mechanical properties and cost for the production of the material. Based on the chemical composition obtained, the casting and subsequent heat treatment of the material was performed. With the produced material, it was machined and mechanical and metallographic tests such as tensile test, hardness, optical test, scanning microscopy and x-ray diffraction were performed. As a result, it was understood that artificial neural networks can be used to assist in the production of an ADI that reaches standardized values and at a considerably lower cost compared to its competitors, with savings of up to 49%. Thus, a smaller number of natural resources can be used achieving the desired mechanical and microstructural properties.

MEMBROS DA BANCA:
Interno - 1934288 - CINTHIA GABRIELY ZIMMER
Interno - 1446153 - DANIELA LUPINACCI VILLANOVA
Externo à Instituição - VINICIUS KARLINSKI DE BARCELLOS - UFRGS